Algorithmic search for baseline minimum aberration designs

• The formulation of the K-aberration criterion for baseline designs is reviewed.
• An efficient incomplete search algorithm to find near optimal baseline designs is proposed.
• Lower bounds for values of K2 and K3 are established.
• A catalogue of optimal (or near optimal) 20-run baseline designs is provided.

This paper reviews the formulation of the K-aberration criterion for baseline two-level designs and the efficient complete search algorithm developed by Mukerjee and Tang (2012). An efficient incomplete search algorithm is proposed that can be used to find near optimal baseline designs in situations where the complete search algorithm is not feasible. Lower bounds for values of K2 and K3 are established. A catalogue of optimal (or near optimal) 20-run baseline designs is provided.